-
Notifications
You must be signed in to change notification settings - Fork 4.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Improve error messages for concurrent CDK #34754
Improve error messages for concurrent CDK #34754
Conversation
The latest updates on your projects. Learn more about Vercel for Git ↗︎ 1 Ignored Deployment
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think this is definitely an improvement, and as such can be merged — I suggest that we figure out how we want folks to contact us, and state that more specifically.
I think issues are fine as we will get a triage process in place. The likelihood that we will be able to help in GH issues is higher than on Slack. Right?
# We do not expect this error to happen. The futures created during concurrent syncs should catch the exception and | ||
# push it to the queue. If this exception occurs, please review the futures and how they handle exceptions. | ||
self._most_recently_seen_exception = RuntimeError( | ||
f"Failed processing a future: {optional_exception}. Please contact the Airbyte team." |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Let's clarify how the users should contact our team. If this situation is a bug and/or requires work from the CDK team, we should ask folks to file an issue on airbytehq/airbyte, right?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I can't confirm as I'm not sure what is our official support process. Maybe @girarda can chime in?
EDIT: I'll merge this change regardless as it is like that elsewhere in the code. Note that this is not because I don't see this as important but I see this as a different issue. I would also like not having specific line of codes share the support process and have a clear page where this is described as this is easier to maintain.
# If we track state value using records cursor field, we can only do that if there is one partition. This is because we save | ||
# the state every time we close a partition. We assume that if there are multiple slices, they need to be providing | ||
# boundaries. There are cases where partitions could not have boundaries: | ||
# * The cursor should be per-partition | ||
# * The stream state is actually the parent stream state | ||
# There might be other cases not listed above. Those are not supported today hence the stream should not use this cursor for | ||
# state management. For the specific user that was affected with this issue, we need to: | ||
# * Fix state tracking (which is currently broken) | ||
# * Make the new version available | ||
# * (Probably) ask the user to reset the stream to avoid data loss |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you for the details here!
What
Adding documentation through comments and improving error messages